As electronic devices become increasingly sophisticated, people are using such devices in new and interesting ways. Some of these devices have adopted voice control where the device can perform various actions in response to a spoken question or instruction. For example, a user might issue an instruction to communicate with a specified recipient. A system receiving information for the instruction can attempt to determine the recipient for the request and cause a connection to be established from the receiving device to a device associated with the recipient. It might be the case, however, that the recipient is associated with multiple devices, and one or more of these devices might be shared with another user. Conventional approaches to maintaining the correct addresses and records for these situations are very resource intensive and can require the knowledge of various different addresses to which a particular device might be registered.